storage and file structure

5/26/2018 Storage and File Structure

1/85

Database System Concepts, 5th Ed.

Silberschatz, Korth and Sudarshan

See www.db-book.comfor conditions on re-use

Storage and File Structure

CS 3040

09/10/2009
http://www.db-book.com/http://www.db-book.com/http://www.db-book.com/http://www.db-book.com/


2/85

Silberschatz, Korth and Sudarshan11.2Database System Concepts - 5thEdition, Oct 23, 2005.

Chapter 11: Storage and File Structure

Overview of Physical Storage Media

Magnetic Disks

RAID

Tertiary Storage

Storage Access

File Organization Organization of Records in Files

Data-Dictionary Storage

Storage Structures for Object-Oriented Databases


3/85


Classification of Physical Storage Media

Speed with which data can be accessed

Cost per unit of data

Reliability

data loss on power failure or system crash

physical failure of the storage device

Can differentiate storage into:

volatile storage: loses contents when power is switched off

non-volatile storage:

Contents persist even when power is switched off.

Includes secondary and tertiary storage, as well as batter-backed up main-memory.


4/85


Physical Storage Media

Cachefastest and most costly form of storage; volatile; managedby the computer system hardware.

Main memory:

fast access (10s to 100s of nanoseconds; 1 nanosecond = 109seconds)

generally too small (or too expensive) to store the entiredatabase

capacities of up to a few Gigabytes widely used currently

Capacities have gone up and per-byte costs havedecreased steadily and rapidly (roughly factor of 2 every 2to 3 years)

Volatilecontents of main memory are usually lost if a powerfailure or system crash occurs.


5/85


Physical Storage Media (Cont.)

Flash memory

Data survives power failure

Data can be written at a location only once, but location can beerased and written to again

Can support only a limited number (10K1M) of write/erasecycles.

Erasing of memory has to be done to an entire bank ofmemory

Reads are roughly as fast as main memory

But writes are slow (few microseconds), erase is slower

Cost per unit of storage roughly similar to main memory Widely used in embedded devices such as digital cameras

Is a type of EEPROM (Electrically Erasable ProgrammableRead-Only Memory)


6/85



Magnetic-disk

Data is stored on spinning disk, and read/written magnetically

Primary medium for the long-term storage of data; typically stores entiredatabase.

Data must be moved from disk to main memory for access, and writtenback for storage

Much slower access than main memory (more on this later)

direct-access possible to read data on disk in any order, unlikemagnetic tape

Capacities range up to roughly 400 GB currently

Much larger capacity and cost/byte than main memory/flash memory

Growing constantly and rapidly with technology improvements (factorof 2 to 3 every 2 years)

Survives power failures and system crashes

disk failure can destroy data, but is rare


7/85Silberschatz, Korth and Sudarshan11.7Database System Concepts - 5thEdition, Oct 23, 2005.


Optical storage

non-volatile, data is read optically from a spinning disk usinga laser

CD-ROM (640 MB) and DVD (4.7 to 17 GB) most popularforms

Write-one, read-many (WORM) optical disks used for archival

storage (CD-R, DVD-R, DVD+R)

Multiple write versions also available (CD-RW, DVD-RW,DVD+RW, and DVD-RAM)

Reads and writes are slower than with magnetic disk

Juke-boxsystems, with large numbers of removable disks, a

few drives, and a mechanism for automatic loading/unloadingof disks available for storing large volumes of data




Tape storage

non-volatile, used primarily for backup (to recover from diskfailure), and for archival data

sequential-accessmuch slower than disk

very high capacity (40 to 300 GB tapes available)

tape can be removed from drive storage costs muchcheaper than disk, but drives are expensive

Tape jukeboxes available for storing massive amounts ofdata

hundreds of terabytes (1 terabyte = 109 bytes) to even apetabyte (1 petabyte = 1012bytes)



Storage Hierarchy



Storage Hierarchy (Cont.)

primary storage: Fastest media but volatile (cache, main

memory).

secondary storage:next level in hierarchy, non-volatile,moderately fast access time

also called on-line storage

E.g. flash memory, magnetic disks

tertiary storage:lowest level in hierarchy, non-volatile, slowaccess time

also called off-line storage

E.g. magnetic tape, optical storage



Magnetic Hard Disk Mechanism

NOTE: Diagram is schematic, and simplifies the structure of actual disk drives



Magnetic Disks

Read-write head

Positioned very close to the platter surface (almost touching it)

Reads or writes magnetically encoded information.

Surface of platter divided into circular tracks

Over 50K-100K tracks per platter on typical hard disks

Each track is divided into sectors.

A sector is the smallest unit of data that can be read or written.

Sector size typically 512 bytes

Typical sectors per track: 500 (on inner tracks) to 1000 (on outer tracks)

To read/write a sector

disk arm swings to position head on right track

platter spins continually; data is read/written as sector passes under head

Head-disk assemblies

multiple disk platters on a single spindle (1 to 5 usually)

one head per platter, mounted on a common arm.

Cylindericonsists of ithtrack of all the platters



Magnetic Disks (Cont.)

Earlier generation disks were susceptible to head-crashes

Surface of earlier generation disks had metal-oxide coatings whichwould disintegrate on head crash and damage all data on disk

Current generation disks are less susceptible to such disastrousfailures, although individual sectors may get corrupted

Disk controllerinterfaces between the computer system and the diskdrive hardware.

accepts high-level commands to read or write a sector initiates actions such as moving the disk arm to the right track and

actually reading or writing the data

Computes and attaches checksumsto each sector to verify thatdata is read back correctly

If data is corrupted, with very high probability stored checksum

wont match recomputed checksum Ensures successful writing by reading back sector after writing it

Performs remapping of bad sectors



Disk Subsystem

Multiple disks connected to a computer system through a controller

Controllers functionality (checksum, bad sector remapping) oftencarried out by individual disks; reduces load on controller

Disk interface standards families ATA(AT adaptor) range of standards

SATA(Serial ATA)

SCSI(Small Computer System Interconnect) range of standards

Several variants of each standard (different speeds and capabilities)



Performance Measures of Disks

Access timethe time it takes from when a read or write request is issued towhen data transfer begins. Consists of:

Seek timetime it takes to reposition the arm over the correct track.

Average seek time is 1/2 the worst case seek time.

Would be 1/3 if all tracks had the same number of sectors, and weignore the time to start and stop arm movement

4 to 10 milliseconds on typical disks

Rotational latencytime it takes for the sector to be accessed to appearunder the head.

Average latency is 1/2 of the worst case latency.

4 to 11 milliseconds on typical disks (5400 to 15000 r.p.m.)

Data-transfer ratethe rate at which data can be retrieved from or stored tothe disk.

25 to 100 MB per second max rate, lower for inner tracks

Multiple disks may share a controller, so rate that controller can handle isalso important

E.g. ATA-5: 66 MB/sec, SATA: 150 MB/sec, Ultra 320 SCSI: 320 MB/s

Fiber Channel (FC2Gb): 256 MB/s



Performance Measures (Cont.)

Mean time to failure (MTTF)the average time the disk isexpected to run continuously without any failure.

Typically 3 to 5 years

Probability of failure of new disks is quite low, corresponding to atheoretical MTTF of 500,000 to 1,200,000 hours for a new disk

E.g., an MTTF of 1,200,000 hours for a new disk means thatgiven 1000 relatively new disks, on an average one will failevery 1200 hours

MTTF decreases as disk ages



Optimization of Disk-Block Access

Blocka contiguous sequence of sectors from a single track

data is transferred between disk and main memory in blocks

sizes range from 512 bytes to several kilobytes

Smaller blocks: more transfers from disk

Larger blocks: more space wasted due to partially filled blocks

Typical block sizes today range from 4 to 16 kilobytes Disk-arm-schedulingalgorithms order pending accesses to tracks so

that disk arm movement is minimized

elevator algorithm: move disk arm in one direction (from outer toinner tracks or vice versa), processing next request in that

direction, till no more requests in that direction, then reversedirection and repeat



Optimization of Disk Block Access (Cont.)

File organizationoptimize block access time by organizing the

blocks to correspond to how data will be accessed E.g. Store related information on the same or nearby cylinders.

Files may get fragmentedover time

E.g. if data is inserted to/deleted from the file

Or free blocks on disk are scattered, and newly created filehas its blocks scattered over the disk

Sequential access to a fragmented file results in increaseddisk arm movement

Some systems have utilities to defragmentthe file system, inorder to speed up file access



Nonvolatile write buffersspeed up disk writes by writing blocks to a non-volatile

RAM buffer immediately Non-volatile RAM: battery backed up RAM or flash memory

Even if power fails, the data is safe and will be written to disk when powerreturns

Controller then writes to disk whenever the disk has no other requests orrequest has been pending for some time

Database operations that require data to be safely stored before continuing cancontinue without waiting for data to be written to disk

Writes can be reordered to minimize disk arm movement

Log diska disk devoted to writing a sequential log of block updates

Used exactly like nonvolatile RAM

Write to log disk is very fast since no seeks are required

No need for special hardware (NV-RAM)

File systems typically reorder writes to disk to improve performance

Journaling file systems write data in safe order to NV-RAM or log disk

Reordering without journaling: risk of corruption of file system data

Optimization of Disk Block Access (Cont.)



RAID

RAID: Redundant Arrays of Independent Disks

disk organization techniques that manage a large numbers of disks,providing a view of a single disk of

high capacityand high speed by using multiple disks in parallel,and

high reliabilityby storing data redundantly, so that data can berecovered even if a disk fails

The chance that some disk out of a set of Ndisks will fail is much higherthan the chance that a specific single disk will fail.

E.g., a system with 100 disks, each with MTTF of 100,000 hours(approx. 11 years), will have a system MTTF of 1000 hours (approx.41 days)

Techniques for using redundancy to avoid data loss are critical withlarge numbers of disks

Originally a cost-effective alternative to large, expensive disks

I in RAID originally stood for ``inexpensive

Today RAIDs are used for their higher reliability and bandwidth.

The I is interpreted as independent



Improvement of Reliability via Redundancy

Redundancystore extra information that can be used to rebuild

information lost in a disk failure E.g., Mirroring(orshadowing)

Duplicate every disk. Logical disk consists of two physical disks.

Every write is carried out on both disks

Reads can take place from either disk

If one disk in a pair fails, data still available in the other

Data loss would occur only if a disk fails, and its mirror disk alsofails before the system is repaired

Probability of combined event is very small

Except for dependent failure modes such as fire or buildingcollapse or electrical power surges

Mean time to data lossdepends on mean time to failure,and mean time to repair

E.g. MTTF of 100,000 hours, mean time to repair of 10 hours givesmean time to data loss of 500*106hours (or 57,000 years) for amirrored pair of disks (ignoring dependent failure modes)



Improvement in Performance via Parallelism

Two main goals of parallelism in a disk system:

1. Load balance multiple small accesses to increase throughput

2. Parallelize large accesses to reduce response time.

Improve transfer rate by striping data across multiple disks.

Bit-level stripingsplit the bits of each byte across multiple disks

In an array of eight disks, write bit iof each byte to disk i.

Each access can read data at eight times the rate of a single disk.

But seek/access time worse than for a single disk

Bit level striping is not used much any more

Block-level stripingwith ndisks, block iof a file goes to disk (imod n) + 1

Requests for different blocks can run in parallel if the blocks reside ondifferent disks

A request for a long sequence of blocks can utilize all disks in parallel



RAID Levels

Schemes to provide redundancy at lower cost by using disk striping

combined with parity bits Different RAID organizations, or RAID levels, have differing cost,

performance and reliability characteristics

RAID Level 1: Mirrored diskswith block striping

Offers best write performance.

Popular for applications such as storing log files in a database system.

RAID Level 0: Block striping; non-redundant.

Used in high-performance applications where data lose is not critical.



RAID Levels (Cont.)

RAID Level 2: Memory-Style Error-Correcting-Codes(ECC) with bit striping.

RAID Level 3: Bit-Interleaved Parity a single parity bit is enough for error correction, not just detection, since

we know which disk has failed

When writing data, corresponding parity bits must also be computedand written to a parity bit disk

To recover data in a damaged disk, compute XOR of bits from otherdisks (including parity bit disk)



RAID Levels (Cont.)

RAID Level 3 (Cont.)

Faster data transfer than with a single disk, but fewer I/Os per secondsince every disk has to participate in every I/O.

Subsumes Level 2 (provides all its benefits, at lower cost).

RAID Level 4: Block-Interleaved Parity; uses block-level striping, and keepsa parity block on a separate disk for corresponding blocks from Notherdisks.

When writing data block, corresponding block of parity bits must also becomputed and written to parity disk

To find value of a damaged block, compute XOR of bits fromcorresponding blocks (including parity block) from other disks.



RAID Levels (Cont.)


Provides higher I/O rates for independent block reads than Level 3 block read goes to a single disk, so blocks stored on different

disks can be read in parallel

Provides high transfer rates for reads of multiple blocks than no-striping

Before writing a block, parity data must be computed

Can be done by using old parity block, old value of current blockand new value of current block (2 block reads + 2 block writes)

Or by recomputing the parity value using the new values of blockscorresponding to the parity block

More efficient for writing large amounts of data sequentially

Parity block becomes a bottleneck for independent block writes sinceevery block write also writes to parity disk



RAID Levels (Cont.)

RAID Level 5: Block-Interleaved Distributed Parity; partitions data and parityamong allN+ 1 disks, rather than storing data in Ndisks and parity in 1 disk.

E.g., with 5 disks, parity block for nth set of blocks is stored on disk (nmod5) + 1, with the data blocks stored on the other 4 disks.


28/85


RAID Levels (Cont.)


Higher I/O rates than Level 4.

Block writes occur in parallel if the blocks and their parityblocks are on different disks.

Subsumes Level 4: provides same benefits, but avoids bottleneckof parity disk.

RAID Level 6: P+Q Redundancyscheme; similar to Level 5, butstores extra redundant information to guard against multiple diskfailures.

Better reliability than Level 5 at a higher cost; not used as widely.


29/85


Choice of RAID Level

Factors in choosing RAID level

Monetary cost Performance: Number of I/O operations per second, and bandwidth

during normal operation

Performance during failure

Performance during rebuildof failed disk

Including time taken to rebuild failed disk

RAID 0 is used only when data safety is not important

E.g. data can be recovered quickly from other sources

Level 2 and 4 never used since they are subsumed by 3 and 5

Level 3 is not used anymore since bit-striping forces single block reads toaccess all disks, wasting disk arm movement, which block striping (level 5)avoids

Level 6 is rarely used since levels 1 and 5 offer adequate safety for almostall applications

So competition is between 1 and 5 only


30/85


Choice of RAID Level (Cont.)

Level 1 provides much better write performance than level 5

Level 5 requires at least 2 block reads and 2 block writes to write a singleblock, whereas Level 1 only requires 2 block writes

Level 1 preferred for high update environments such as log disks

Level 1 had higher storage cost than level 5

disk drive capacities increasing rapidly (50%/year) whereas disk accesstimes have decreased much less (x 3 in 10 years)

I/O requirements have increased greatly, e.g. for Web servers

When enough disks have been bought to satisfy required rate of I/O, theyoften have spare storage capacity

so there is often no extra monetary cost for Level 1!

Level 5 is preferred for applications with low update rate,

and large amounts of data

Level 1 is preferred for all other applications


31/85


Hardware Issues

Software RAID: RAID implementations done entirely in software, with

no special hardware support

Hardware RAID: RAID implementations with special hardware

Use non-volatile RAM to record writes that are being executed

Beware: power failure during write can result in corrupted disk

E.g. failure after writing one block but before writing the second

in a mirrored system Such corrupted data must be detected when power is restored

Recovery from corruption is similar to recovery from faileddisk

NV-RAM helps to efficiently detected potentially corruptedblocks

Otherwise all blocks of disk must be read and comparedwith mirror/parity block


32/85


Hardware Issues (Cont.)

Hot swapping: replacement of disk while system is running,

without power down Supported by some hardware RAID systems,

reduces time to recovery, and improves availability greatly

Many systems maintain spare diskswhich are kept online, andused as replacements for failed disks immediately on detection

of failure Reduces time to recovery greatly

Many hardware RAID systems ensure that a single point offailure will not stop the functioning of the system by using

Redundant power supplies with battery backup

Multiple controllers and multiple interconnections to guardagainst controller/interconnection failures


33/85


Optical Disks

Compact disk-read only memory (CD-ROM)

Removable disks, 640 MB per disk Seek time about 100 msec (optical read head is heavier and slower)

Higher latency (3000 RPM) and lower data-transfer rates (3-6 MB/s)compared to magnetic disks

Digital Video Disk (DVD)

DVD-5 holds 4.7 GB , and DVD-9 holds 8.5 GB

DVD-10 and DVD-18 are double sided formats with capacities of 9.4GB and 17 GB

Slow seek time, for same reasons as CD-ROM

Record once versions (CD-R and DVD-R) are popular

data can only be written once, and cannot be erased.

high capacity and long lifetime; used for archival storage Multi-write versions (CD-RW, DVD-RW, DVD+RW and DVD-RAM)

also available


34/85


Magnetic Tapes

Hold large volumes of data and provide high transfer rates

Few GB for DAT (Digital Audio Tape) format, 10-40 GB with DLT(Digital Linear Tape) format, 100 GB+ with Ultrium format, and 330GB with Ampex helical scan format

Transfer rates from few to 10s of MB/s

Currently the cheapest storage medium

Tapes are cheap, but cost of drives is very high

Very slow access time in comparison to magnetic disks and opticaldisks

limited to sequential access.

Some formats (Accelis) provide faster seek (10s of seconds) atcost of lower capacity

Used mainly for backup, for storage of infrequently used information,

and as an off-line medium for transferring information from one systemto another.

Tape jukeboxes used for very large capacity storage

(terabyte (1012 bytes) to petabye (1015 bytes)


35/85


Storage Access

A database file is partitioned into fixed-length storage units called

blocks. Blocks are units of both storage allocation and datatransfer.

Database system seeks to minimize the number of block transfersbetween the disk and memory. We can reduce the number ofdisk accesses by keeping as many blocks as possible in mainmemory.

Bufferportion of main memory available to store copies of diskblocks.

Buffer managersubsystem responsible for allocating bufferspace in main memory.


36/85


Buffer Manager

Programs call on the buffer manager when they need a block

from disk.1. If the block is already in the buffer, buffer manager returns

the address of the block in main memory

2. If the block is not in the buffer, the buffer manager

1. Allocates space in the buffer for the block

1. Replacing (throwing out) some other block, if required,to make space for the new block.

2. Replaced block written back to disk only if it wasmodified since the most recent time that it was writtento/fetched from the disk.

2. Reads the block from the disk to the buffer, and returnsthe address of the block in main memory to requester.


37/85


Buffer-Replacement Policies

Most operating systems replace the block least recently used

(LRU strategy) Idea behind LRUuse past pattern of block references as a

predictor of future references

Queries have well-defined access patterns (such as sequentialscans), and a database system can use the information in a users

query to predict future references

LRU can be a bad strategy for certain access patterns involvingrepeated scans of data

For example: when computing the join of 2 relations r and sby a nested loopsfor each tuple trof rdo

for each tuple tsof sdoif the tuples trand tsmatch

Mixed strategy with hints on replacement strategy providedby the query optimizer is preferable

B ff R l P li i (C )


38/85


Buffer-Replacement Policies (Cont.)

Pinned blockmemory block that is not allowed to be written

back to disk. Toss-immediatestrategyfrees the space occupied by a block

as soon as the final tuple of that block has been processed

Most recently used (MRU) strategy system must pin the blockcurrently being processed. After the final tuple of that block hasbeen processed, the block is unpinned, and it becomes the mostrecently used block.

Buffer manager can use statistical information regarding theprobability that a request will reference a particular relation

E.g., the data dictionary is frequently accessed. Heuristic:keep data-dictionary blocks in main memory buffer

Buffer managers also support forced outputof blocks for thepurpose of recovery (more in Chapter 17)

Fil O i i


39/85


File Organization

The database is stored as a collection of files. Each file is a

sequence of records. A record is a sequence of fields.

One approach:

assume record size is fixed

each file has records of one particular type only

different files are used for different relationsThis case is easiest to implement; will consider variable lengthrecords later.

Fi d L th R d


40/85


Fixed-Length Records

Simple approach:

Store record istarting from byte n (i

1), where n is the size ofeach record.

Record access is simple but records may cross blocks

Modification: do not allow records to cross block boundaries

Deletion of record i:alternatives:

move records i+ 1, . . ., nto i, . . . , n 1

move record n to i

do not move records, butlink all free records on afree list

F Li t


41/85


Free Lists

Store the address of the first deleted record in the file header.

Use this first record to store the address of the second deleted record,and so on

Can think of these stored addresses as pointerssince they point tothe location of a record.

More space efficient representation: reuse space for normal attributes

of free records to store pointers. (No pointers stored in in-use records.)

V i bl L th R d


42/85


Variable-Length Records

Variable-length records arise in database systems in several

ways:

Storage of multiple record types in a file.

Record types that allow variable lengths for one or morefields.

Record types that allow repeating fields (used in some

older data models).

V i bl L th R d Sl tt d P St t


43/85


Variable-Length Records: Slotted Page Structure

Slotted pageheader contains: number of record entries

end of free space in the block

location and size of each record

Records can be moved around within a page to keep them contiguouswith no empty space between them; entry in the header must beupdated.

Pointers should not point directly to record instead they shouldpoint to the entry for the record in header.

O i ti f R d i Fil


44/85


Organization of Records in Files

Heapa record can be placed anywhere in the file where there

is space Sequentialstore records in sequential order, based on the

value of the search key of each record

Hashinga hash function computed on some attribute of eachrecord; the result specifies in which block of the file the record

should be placed Records of each relation may be stored in a separate file. In a

multitable clustering file organization records of severaldifferent relations can be stored in the same file

Motivation: store related records on the same block to

minimize I/O

S ti l Fil O i ti


45/85


Sequential File Organization

Suitable for applications that require sequential processing of

the entire file The records in the file are ordered by a search-key

Seq ential File Organi ation (Cont )


46/85


Sequential File Organization (Cont.)

Deletionuse pointer chains

Insertionlocate the position where the record is to be inserted if there is free space insert there

if no free space, insert the record in an overflow block

In either case, pointer chain must be updated

Need to reorganize the filefrom time to time to restoresequential order

Multitable Clustering File Organization


47/85


Multitable Clustering File Organization

Store several relations in one file using a multitable clustering

file organization

M ltit bl Cl t i Fil O i ti ( t )


48/85


Multitable Clustering File Organization (cont.)

Multitable clustering organization of customer and depositor:

good for queries involving depositor customer, and for queriesinvolving one single customer and his accounts

bad for queries involving only customer

results in variable size records

Can add pointer chains to link records of a particular relation

Data Dictionary Storage


49/85


Data Dictionary Storage

Information about relations

names of relations

names and types of attributes of each relation

names and definitions of views

integrity constraints User and accounting information, including passwords

Statistical and descriptive data

number of tuples in each relation

Physical file organization information

How relation is stored (sequential/hash/)

Physical location of relation

Information about indices (Chapter 12)

Data dictionary(also called system catalog) stores metadata;

that is, data about data, such as

Data Dictionary Storage (Cont )


50/85


Data Dictionary Storage (Cont.)

Catalog structure

Relational representation on disk specialized data structures designed for efficient access, in

memory

A possible catalog representation:

Relation_metadata = (relation_name, number_of_attributes,storage_organization, location)

Attribute_metadata = (attribute_name, relation_name, domain_type,position, length)

User_metadata = (user_name, encrypted_password, group)Index_metadata = (index_name, relation_name, index_type,

index_attributes)View_metadata = (view_name, definition)


51/85

Database System Concepts, 5th Ed.

Silberschatz, Korth and SudarshanSee www.db-book.comfor conditions on re-use

End of Chapter 11

Record Representation
http://www.db-book.com/http://www.db-book.com/http://www.db-book.com/http://www.db-book.com/


52/85


Record Representation

Records with fixed length fields are easy to represent

Similar to records (structs) in programming languages Extensions to represent null values

E.g. a bitmap indicating which attributes are null

Variable length fields can be represented by a pair(offset,length)

where offset is the location within the record and length is field length.

All fields start at predefined location, but extra indirection requiredfor variable length fields

Example record structure of accountrecord

account_number

branch_name

balance

PerryridgeA-102 40010

File Containing account Records


53/85


File Containing account Records

File of Figure 11.6, with Record 2 Deleted and


54/85


g ,All Records Moved

File of Figure 11.6, With Record 2 deleted and


55/85


Final Record Moved

y e- r ng epresen a on o ar a e- engR d


56/85


Records

Clustering File Structure


57/85



Clustering File Structure With Pointer Chains


58/85


Clustering File Structure With Pointer Chains

The deposi tor Relation


59/85


The deposi torRelation

The customer Relation


60/85


The customer Relation



61/85




62/85


Figure 11 4


63/85


Figure 11.4

Figure 11 7


64/85


Figure 11.7

Figure 11.8


65/85


Figure 11.8

Figure 11.100


66/85


Figure 11.100

Figure 11.20


67/85


Figure 11.20

Byte-String Representation of Variable-Length Records


68/85


y g p g

Byte string representationAttach an end-of-record() control character to the end of each recordDifficulty with deletion

Difficulty with growth

Fixed-Length Representation


69/85


Fixed Length Representation

Use one or more fixed length records:

reserved space pointers

Reserved spacecan use fixed-length records of a knownmaximum length; unused space in shorter records filled with a nullor end-of-record symbol.

Pointer Method


70/85


Pointer Method

Pointer method A variable-length record is represented by a list of fixed-length

records, chained together via pointers.

Can be used even if the maximum record length is not known

Pointer Method (Cont )


71/85


Pointer Method (Cont.)

Disadvantage to pointer structure; space is wasted in allrecords except the first in a a chain.

Solution is to allow two kinds of block in file:

Anchor blockcontains the first records of chain

Overflow blockcontains records other than those thatare the first records of chairs.

Mapping of Objects to Files


72/85


pp g Obj

Mapping objects to files is similar to mapping tuples to files in arelational system; object data can be stored using file structures.

Objects in O-O databases may lack uniformity and may be very large;such objects have to managed differently from records in a relationalsystem.

Set fields with a small number of elements may be implementedusing data structures such as linked lists.

Set fields with a larger number of elements may be implemented asseparate relations in the database.

Set fields can also be eliminated at the storage level bynormalization.

Similar to conversion of multivalued attributes of E-R diagrams torelations

Mapping of Objects to Files (Cont.)


73/85


pp g j ( )

Objects are identified by an object identifier (OID); the storage systemneeds a mechanism to locate an object given its OID (this action iscalled dereferencing).

logical identifiersdo not directly specify an objects physicallocation; must maintain an index that maps an OID to the objects

actual location.

physical identifiersencode the location of the object so the

object can be found directly. Physical OIDs typically have thefollowing parts:

1. a volume or file identifier

2. a page identifier within the volume or file

3. an offset within the page

Management of Persistent Pointers


74/85


g

Physical OIDs may be a unique identifier. This identifier isstored in the object also and is used to detect references viadangling pointers.

Management of Persistent Pointers(Cont )


75/85


(Cont.)

Implement persistent pointers using OIDs; persistent pointers aresubstantially longer than are in-memory pointers

Pointer swizzling cuts down on cost of locating persistent objectsalready in-memory.

Software swizzling (swizzling on pointer deference)

When a persistent pointer is first dereferenced, the pointer is

swizzled(replaced by an in-memory pointer) after the object islocated in memory.

Subsequent dereferences of of the same pointer become cheap.

The physical location of an object in memory must not change ifswizzled pointers pont to it; the solution is to pin pages in memory

When an object is written back to disk, any swizzled pointers itcontains need to be unswizzled.

Hardware Swizzling


76/85


g

With hardware swizzling, persistent pointers in objects need thesame amount of space as in-memory pointers extra storageexternal to the object is used to store rest of pointer information.

Uses virtual memory translation mechanism to efficiently andtransparently convert between persistent pointers and in-memorypointers.

All persistent pointers in a page are swizzled when the page is

first read in.

thus programmers have to work with just one type of pointer,i.e., in-memory pointer.

some of the swizzled pointers may point to virtual memoryaddresses that are currently not allocated any real memory

(and do not contain valid data)

Hardware Swizzling


77/85


g

Persistent pointer is conceptually split into two parts: a page identifier,and an offset within the page.

The page identifier in a pointer is a short indirect pointer: Eachpage has a translation table that provides a mapping from theshort page identifiers to full database page identifiers.

Translation table for a page is small (at most 1024 pointers in a4096 byte page with 4 byte pointer)

Multiple pointers in page to the same page share same entry inthe translation table.

Hardware Swizzling (Cont.)


78/85


g ( )

Page image before swizzling (page located on disk)



79/85


g ( )

When system loads a page into memory the persistent pointers in the pageare swizzledas described below

1. Persistent pointers in each object in the page are located using objecttype information

2. For each persistent pointer (pi, oi) find its full page ID Pi

1. If Pidoes not already have a virtual memory page allocated to it,

allocate a virtual memory page to Pi and read-protect the page Note: there need not be any physical space (whether in memory

or on disk swap-space) allocated for the virtual memory page atthis point. Space can be allocated later if (and when) Pi isaccessed. In this case read-protection is not required.

Accessing a memory location in the page in the will result in asegmentation violation, which is handled as described later

2. Let vibe the virtual page allocated to Pi(either earlier or above)3. Replace (pi, oi) by (vi, oi)

3. Replace each entry (pi, Pi) in the translation table, by (vi, Pi)



80/85


g ( )

When an in-memory pointer is dereferenced, if the operatingsystem detects the page it points to has not yet been allocatedstorage, or is read-protected, a segmentation violationoccurs.

The mmap() call in Unix is used to specify a function to be invokedon segmentation violation

The function does the following when it is invoked

1.Allocate storage (swap-space) for the page containing thereferenced address, if storage has not been allocated earlier.Turn off read-protection

2. Read in the page from disk

3. Perform pointer swizzling for each persistent pointer in thepage, as described earlier



81/85


g

Page with short page identifier 2395 was allocated address 5001.Observe change in pointers and translation table.

Page with short page identifier 4867 has been allocated address4867. Nochange in pointer and translation table.

Page image after swizzling



82/85


After swizzling, all short page identifiers point to virtual memory addressesallocated for the corresponding pages

functions accessing the objects are not even aware that it haspersistent pointers, and do not need to be changed in any way!

can reuse existing code and libraries that use in-memory pointers

After this, the pointer dereference that triggered the swizzling can continue

Optimizations: If all pages are allocated the same address as in the short page

identifier, no changes required in the page!

No need for deswizzling swizzled page can be saved as-is to disk

A set of pages (segment) can share one translation table. Pages can

still be swizzled as and when fetched (old copy of translation table isneeded).

A process should not access more pages than size of virtual memory reuse of virtual memory addresses for other pages is expensive

Disk versus Memory Structure of Objects


83/85


The format in which objects are stored in memory may be different fromthe formal in which they are stored on disk in the database. Reasonsare:

software swizzlingstructure of persistent and in-memory pointersare different

database accessible from different machines, with different datarepresentations

Make the physical representation of objects in the databaseindependent of the machine and the compiler.

Can transparently convert from disk representation to form requiredon the specific machine, language, and compiler, when the object(or page) is brought into memory.

Large Objects


84/85


Large objects : binary large objects (blobs) and character largeobjects (clobs)

Examples include:

text documents

graphical data such as images and computer aided designsaudio and video data

Large objects may need to be stored in a contiguous sequence ofbytes when brought into memory.

If an object is bigger than a page, contiguous pages of the bufferpool must be allocated to store it.

May be preferable to disallow direct access to data, and only allow

access through a file-system-like API, to remove need forcontiguous storage.

Modifying Large Objects


85/85

If the application requires insert/delete of bytes from specified regions ofan object:

B+-tree file organization (described later in Chapter 12) can bemodified to represent large objects

Each leaf page of the tree stores between half and 1 page worth ofdata from the object

Special-purpose application programs outside the database are used tomanipulate large objects:

Text data treated as a byte string manipulated by editors andformatters.

Graphical data and audio/video data is typically created and displayedby separate application

checkout/checkinmethod for concurrency control and creation ofversions

storage and file structure

Documents

unit of storage

volatile storage

databases5262018 storage

memory5262018 storage

storage device

longterm storage of

database system concepts

flash memory data